tbh's School-by-School Takeover: How Nikita Bier Engineered Viral Critical Mass in 9 Weeks

tbh went from zero to the #1 app in the United States in roughly six days — and sold to Facebook nine weeks later — with a team of four people and almost no money. The engine was a precise, school-by-school seeding playbook: saturate one school with enough coordinated adoption to light the social flywheel, then let the positivity mechanic (you can only send compliments, never insults) drive explosive peer-to-peer spread. One school sent 450,000 messages in its first seven days. At peak, tbh was generating 360,000 installs per day.

The Problem Anonymous Apps Always Had

By 2017 the app store signal was unmistakable: Sarahah, an anonymous-messaging app entirely in Arabic, had become the #1 app in the United States. As Nikita Bier put it, "That was one of the strongest signals you could ever have that people want something." Teens were desperate for a vehicle to receive honest, anonymous feedback from their peers — but every anonymous app that had gone viral before had curdled into bullying and self-harm, generating the kind of news stories Bier had sworn he'd never be part of.

The insight that broke the logjam came from watching teens on Snapchat playing a game they called tbh ("to be honest"): they would post a collage of friend emojis and invite replies about who had the best smile, who was most likely to be president. The loop was entirely positive. Bier's realization: "What if instead of actually typing what you wanted to say about somebody, you just answered polls and we authored those polls so that we ensured everything would always be positive?" By controlling the content format — multiple-choice compliment polls, not open text — tbh could deliver the anonymous validation teens craved while making bullying structurally impossible.

The Seeding Playbook: Eliminate Every Confounding Variable

After 14 failed apps over four to five years, Bier's team had developed a precise method for seeding a new product into a school. The goal was not just to get downloads — it was to get enough simultaneous adoption that the social network would actually function. "You need to get enough intensity of adoption and density for a social network to start to get the flywheel spinning."

The mechanics: run targeted ads at the specific school, and build a dedicated Instagram account that followed every student who had their school abbreviation in their bio (high schoolers consistently tag themselves this way). Accept every followback. Saturate the school with enough marketing touches — Bier's rule of thumb was that a user needs to see a message roughly three times before downloading — so that a meaningful slice of the school adopted synchronously on the same day. The key discipline: this was how they tested apps, not how they expected to grow them. Once the initial density was achieved, the app had to spread by itself.

For tbh's launch, they chose a school in Georgia with the earliest start date in the United States — they needed signal fast, given the company's financial state. The result: "I think it spread to 40% of the school [that] downloaded it in the first 24 hours and it rapidly spread to the neighboring schools."

The Positivity Mechanic as Viral Fuel

The product design did the rest. Because tbh only allowed positive anonymous polls, users who received a compliment notification felt good — and wanted to send more. Day-one messaging volume at one school reached 60 messages per user, against a typical new-app benchmark of "three or four or something." One school sent 450,000 messages in its first seven days of adoption.

The notification architecture was equally deliberate. When you received a poll result — say, "Someone thinks you have the best smile" — you could see which grade the person was in, giving just enough context to intrigue without revealing identity. That friction-with-curiosity loop drove people back into the app to send more polls, hoping to identify their admirer. Volume compounded.

Bier also built a fairness system into the poll-selection algorithm: "We put your name in polls at a higher frequency if you weren't being voted on recently. So we wanted to spread the love in every way possible." This prevented the social stratification that typically makes teen social apps feel exclusive and therefore unsafe for lower-status users.

State-by-State Geo-fencing as Growth Lever

As servers began to buckle, Bier made a counterintuitive call: geo-fence the app to slow its own growth. Rather than let it spread nationally before infrastructure could handle the load, the team rolled out state by state. This controlled-burn approach meant they could keep the product fast and reliable in each market — essential for an app whose entire value proposition was the rush of instant notification — while rebuilding their AWS stack (their bill was approaching $120,000 with only $150,000 in the bank).

The state sequencing also created a cinematic sales moment. When meeting a potential acquirer in Los Angeles, Bier showed the app's real-time install map — built by their CTO, a published mapping author — just hours after California had been unlocked. He zoomed in on the block they were sitting on. The entire street was lit up with installs happening in real time, radiating outward from the meeting room. "That's what got the ball rolling."

The Outcome: #1 in Six Days, Acquired in Nine Weeks

At peak, tbh was recording 360,000 installs per day — enough to be the #1 free app in the United States. The number-one threshold at the time was roughly 80,000–100,000 daily installs; tbh was running at more than triple that. Nine weeks after launch, Bier sold the company to Facebook for over $30 million. The entire team was four people: two engineers, one designer, and Bier.

The Lesson

Virality is not a feature you add — it's a system you design from the ground up. Bier's playbook has three interlocking parts: (1) a product mechanic that makes users feel something and want to tell their peers (the positivity loop); (2) a seeding method that gets enough simultaneous adoption to make the social network functional before it has to grow on its own; and (3) the discipline to treat each school or cohort as an isolated test — eliminate all confounding variables so you know, with certainty, whether the product has legs. "You never want to walk away from an experiment and say, well, maybe the execution was bad." When the product is right and the seeding is right, growth is not uncertain. It is inevitable.

Challenge

After 14 failed apps and nearly no money left, Bier needed tbh to achieve genuine social-network density fast enough to trigger organic spread — in a category (anonymous teen social) with a well-documented history of bullying, regulatory scrutiny, and churn. The dual problem: how to engineer the initial critical mass, and how to design the product so the positivity mechanic would sustain virality without needing a content moderation army.

Approach

Bier seeded tbh school-by-school using targeted ads and Instagram follows of students who tagged their school in their bios, getting 40% of a school to download on day one. The product was designed so that only positive anonymous compliment polls were possible, generating 60 messages/user on day one vs. a typical 3–4, and keeping press coverage clean. A state-by-state geo-fence controlled growth to match server capacity while preserving the fast, high-quality experience the viral loop depended on.

Results

• School adoption in first 24 hours (launch school, Georgia): ~40% of the school downloaded the app
• Messages sent in first 7 days at one school: 450,000
• Day-1 messages per user vs. typical new app benchmark: 60 vs. 3–4
• Peak daily install rate: 360,000 installs/day
• Time to #1 in the US App Store: ~6 days from launch
• Time from launch to acquisition: 9 weeks
• Acquisition price: Over $30 million (to Facebook)
• Team size at acquisition: 4 people (2 engineers, 1 designer, Nikita Bier)
• Invitations per user decline by age: ~20% per year from age 13 to 18

Sources

• Lenny's Podcast — Nikita Bier (tbh, Gas)

The full record sits in the studio register.

Related

Part of the Referral growth pillar. See also Dropbox's Double-Sided Referral Program, PayPal's Paid Referral Blitz, Tesla's Tiered Referral Program.